From what I have learned on this board . It was wrong of ati to use quack and they had to have thier name tarnished on all the websites. But with nvidia doing the same thing with 3dmark for the second time (disabling splash screens to boost thier score) Its alright cause its the benchmark that shouldn't allow cheating ? heh...
FutureMark will not be accepting NON "WHQL" driver
- Thread starter T2k
- Start date
pocketmoon_
Newcomer
LeStoffer said:
I honestly believe that the driver engineers have a lot of power to harness with the NV30 architecture, and it's taking time.
And, Joe,I don't see what at all oxymoric about 'Official Developer' release. Redundacy perhaps - all Developer releases are bound to be Official after all.
pocketmoon_ said:
Agreed. nVidia driver engineers never had a challenge similar to the NV30.
However, I'd say the reasons for this are multiple.
First, the NV30 is very different from the NV20/NV10/...
Secondly, it got some potential complex FP16/FP32 optimizations.
Thirdly, the NV30 got a lot of bugged silicon.
One and two will certainly allow big performance increases. But the question is whether three could guarantee it never becomes so great.
It'll become more acceptable, at best, IMO. I doubt it'll ever beat the R350.
Uttar
Joe DeFuria
Legend
pocketmoon_ said:
No. most (if not all) "developer" releases are not drivers officially available to the public.
A developer release is "unofficial". Any one specific developer driver build might introduce more problems than they are intended to fix.
Joe DeFuria said:Evildeus said:I think that sums everything
Yup. Very hypocritical of non-beta members to state how "bad" it is for the industry to have synthetic benchmarks, and then produce drivers that they publically admit are optimized for it, and "request" that web-sites limit testing to use those drivers.
Joe I think you are a smart guy so I am not trying to point out the obvious, but ATI's drivers are optimized for 3dmark too, Nvidia just admitted it to prove a point, what point? Well obviously to show why they think it is bad and how easily it can be manipulated, they could have easily hidden what they were doing if they wanted to.
bridpop said:
LOL that was awesome, and that is exactly how I feel reading most of this stuff, I think there is still a backlash against Nvidia because they were the dominant player, and all the ATI fans felt grumpy because there favorite company wasn't competing very well, so now they evangelize and actually have something to talk about too. Because ATI has delivered, the graphics card market is healthy on both sides right now, and that is good for everyone.
WTF is all this talk of backlashes against NVIDIA, or ATI? This is very simple, look at what Futuremark states:
Thats what this is about. Nobody is talking about optimistations being a bad thing, what Futuremark is talking about is whether these opimisations come at the cost of IQ, which is not a good thing, is it?
They appear to be investigating it and should they not find any issues then they will most likely be re-enabled in the database, if they do find something then they will probably remain excluded.
Oh, and FYI, Beyond3D are now a full member of the Beta group.
Thats what this is about. Nobody is talking about optimistations being a bad thing, what Futuremark is talking about is whether these opimisations come at the cost of IQ, which is not a good thing, is it?
They appear to be investigating it and should they not find any issues then they will most likely be re-enabled in the database, if they do find something then they will probably remain excluded.
Oh, and FYI, Beyond3D are now a full member of the Beta group.
pocketmoon_
Newcomer
Joe DeFuria said:
I disagree. How can a developer release be 'unofficial' ? Perhaps if a developer is hitting a known bug and NV release a specially fixed version to that developer, but 43.40 is on general release to all registered developers.
It is these Developer releases that go on to become release candidates and official public drivers.
And as for the "We are investigating the drivers and their effect on 3DMark03 - both performance and the rendering quality (ie. image quality)." blurb, why don't they find something to complain about FIRST rather than this guilty until proved innocent nonsense. It's undemocratic!
Joe DeFuria
Legend
pocketmoon_ said:
Because it is not officially available to the public.
I dopn't understand the resistance to this simple concept. If the drivers are not deemed suitable enough by nVidia for consumer consumption for whatever reason, then it is not an official driver release.
It is not a general or even beta release to consumers.
I thought the "something to complain about" has already been clearly established? (Image quality of lighting in game test 4.)
Joe DeFuria
Legend
Sxotty said:
Thanks...don't let the secret out though...
I don't care how "optimized" something is, as long as it does not impact image quality. That's the point. If you can find an ATI driver where it doesn't represent what the reference image does, then that driver shouldn't be used. The particular nVidia drivers in question have an impact on image quality.
If nVidia produces some drivers that can faithfully create the image quality they are suppossed to, and at the same time increase performance, that doesn't go to prove their that benchmarks can be manipulated.
That goes to prove their past drivers were horribly inefficient. (Something they have been basically claiming regardless, btw.)
If nVidia can only increase performance by not creating the reference image quality, then that only goes to prove that anyone can cheat by taking shortcuts that degrade image quality. Not much of a revelation there.
Joe DeFuria
Legend
Evildeus said:Can't be proven and still under investigation
If you are really going to be that absurd, the analogy to the U.S. legal system is the following:
There is in fact sufficient evidence to warrant a trial. In the best interests of the community, the drivers in question are being held without bail, and the ultimate fate of the guilt / innocence is reserved for the completion of the investigation and trial.
Drivers provided to developers are on a official, exclusive and NDA basis with registered developers. These are not drivers to be provided on a official basis to the public. As such, they are both "official" as well as "unofficial". I do not see anywhere in this thread about NVIDIA drivers given specifically to review sites for testing/benchmarking 3DMark03 and that such drivers are not to be provided officially... that will be another can of worms which have been discussed before. And is not what I think pocketmoon_ is talking about - he is, I believe, speaking entirely from the point of view of a "developer".Joe DeFuria said:
How's that, Joe? 8)
Well, when something is proven it means that it has been judged, otherwise it's still under investigation. They are still in the wondering period, not in proven period.
You are the type of guy to say: there's a trial, it's because there's something so there's some guilt...
You are the type of guy to say: there's a trial, it's because there's something so there's some guilt...
Joe DeFuria said:Evildeus said:Can't be proven and still under investigation
If you are really going to be that absurd, the analogy to the U.S. legal system is the following:
There is in fact sufficient evidence to warrant a trial. In the best interests of the community, the drivers in question are being held without bail, and the ultimate fate of the guilt / innocence is reserved for the completion of the investigation and trial.
